New machine learning-based method helps detect atrial fibrillation drivers

Researchers from Skoltech and their US colleagues have designed a brand new machine learning-based method for detecting atrial fibrillation drivers, small patches of the center muscle which are hypothesized to trigger this commonest sort of cardiac arrhythmia. This method might result in extra environment friendly focused medical interventions to deal with the situation that’s estimated to have an effect on greater than 33 million folks worldwide, in response to the American Heart Association. The latest paper was printed within the journal Circulation: Arrhythmia and Electrophysiology.

The mechanism behind atrial fibrillation (AF), a sort of irregular coronary heart rhythm that’s related to elevated threat of coronary heart failure and stroke, is but unclear. Research suggests it might be induced and maintained by what’s referred to as reentrant AF drivers, extremely localized sources of repetitive rotational exercise that result in irregular coronary heart rhythm. These drivers may be burnt through a surgical process, which might mitigate the situation and even restore the traditional functioning of the center.

To find these reentrant AF drivers for subsequent destruction, medical doctors use multi-electrode mapping, a method that permits them to document a number of electrograms inside the center (that is accomplished with a catheter) and construct a map {of electrical} exercise inside the atria. However, medical functions of this method usually produce a number of false negatives, when an present AF driver just isn’t discovered, and false positives, when a driver is detected the place there actually is none.

Recently, researchers have tapped machine studying algorithms for the duty of deciphering ECGs to search for atrial fibrillation; nonetheless, these algorithms require labeled information with the true location of the motive force, and the accuracy of multi-electrode mapping is inadequate. The authors of the brand new examine, co-led by Dmitry Dylov from the Skoltech Center of Computational and Data-Intensive Science and Engineering (CDISE) and Vadim Fedorov from the Ohio State University, used high-resolution near-infrared optical mapping (NIOM) to find AF drivers and caught with it as a reference for coaching.

NIOM is predicated on well-penetrating infrared optical alerts and due to this fact can document {the electrical} exercise from inside the coronary heart muscle, whereas typical medical electrodes can solely measure the alerts on the floor. Add to this trait the superb optical decision, and the optical mapping turns into a no brainer modality if you wish to visualize and perceive {the electrical} sign propagation by the center tissue.”

Dmitry Dylov, Skoltech Center of Computational and Data-Intensive Science and Engineering (CDISE)

The group examined their method on eleven explanted human hearts, all donated posthumously for analysis functions. Researchers carried out the simultaneous optical and multi-electrode mapping of AF episodes induced within the hearts. ML mannequin can certainly effectively interpret electrograms from multielectrode mapping to find AF drivers, with an accuracy of as much as 81%. They imagine that bigger coaching datasets, validated by NIOM, can enhance machine learning-based algorithms sufficient for them to grow to be complementary instruments in medical observe.

“The dataset of recording from 11 human hearts is both extremely priceless and too small. We realized that clinical translation would require a much larger sample size for representative sampling, yet we had to make sure we extracted every piece of available information from the still-beating explanted human hearts. Dedication and scrutiny of two of our PhD students must be acknowledged here: Sasha Zolotarev spent several months on the academic mobility trip to Fedorov’s lab understanding the specifics of the imaging workflow and present the pilot study at the HRS conference – the biggest arrhythmology meeting in the world, and Katya Ivanova partook in the frequency and visualization analysis from within the walls of Skoltech. These two young researchers have squeezed out everything one possibly could, to train the machine learning model using optical measurements,” Dylov notes.


Journal reference:

Zolotarev, A.M., et al. (2020) Optical Mapping-Validated Machine Learning Improves Atrial Fibrillation Driver Detection by Multi-Electrode Mapping. Circulation: Arrhythmia and Electrophysiology.


Please enter your comment!
Please enter your name here